基于视觉的印刷品缺陷检测技术

Research on the Technology of Detecting Printing Defects Based on Computer Vision

为满足印刷品缺陷检测的准确度需求,设计了一种基于计算机视觉的分布式检测系统。该系统主要包括图像采集模块、图像处理模块、输入输出模块。研究了图像预处理、特征提取、图像配准、图像匹配和缺陷检测的相关算法。实验数据表明:该系统和采用的图像算法能精确计算缺陷的各项特征参数并判断类型,达到了印刷品缺陷检测的要求。

According to the accuracy requirement of detecting printing defects,this paper designed a distributed detection system based on computer vision. This system was mainly composed of image acquisition module,image processing module,input and output module. At the same time,this paper studied some related algorithms of image prepossessing,feature extraction,image registration,image matching and defect detection. The experiment results clearly show that this system and those image algorithms adopted can accurately calculate various characteristic parameters and determine types of different defects and can meet the requirement of the inspection of printing flaws.